The development of malignancy in the human breast is a poorly understood process. It is clear, however, that the occurrence of in situ cancers, particularly ductal carcinoma in situ (DCIS), is an important step in the evaluation of some invasive breast cancers. The investigators propose to elucidate t he occurrence, biological features, and etiology of in situ and invasive breast cancers, with a particular focus on ductal carcinoma in situ (DCIS) . They will identify 560 new cases of in situ breast cancer diagnosed over a four-year period in New cases of in situ breast cancer diagnosed over a four-year period in New Hampshire and Vermont, using mammography cell tumor registries in these two states, review of all breast biology reports, and link to the state cancer registries. Of these cases, they expect to enroll 412 women of age 75 or less and to classify their biopsies into histopathological categories of DCIS, using a standard format. Using population data sources, they also will enroll a random sample of 412 women with invasive breast cancer. They will obtain tissue blocks for all 824 invasive and in situ cases for immunohistochemical staining to measure HER-2/neu, p53, estrogen receptor (ER) and progesterone receptor (PR) protein expression. Blocks will be analyzed, using PCR techniques, for the alteration of cancer-related genes p16 and p15 on chromosome 9p, and for extent and frequency of loss of heterozygosity on chromosome 11. As a control, they will identify 412 women without breast malignancy, using the same network resources. By telephone questionnaire, they will assess demographic characteristics, hormones use, and reproductive and lifestyle risk factors for cases and controls. Statistical analyses will assess: (1) the age-specific incidence of in situ breast cancer and its cell types; (2) the prevalence of immunohistochemical and molecular biological markers among in situ and invasive cancers; (3) risk factors for DCIS and its subtypes of (e.g., with or without comedo features: (4) possible associations, within the entire spectrum of breast neoplasia, between risk factors and (a) HER- 2/neu, p53, ER and PR expression,and (b) chromosomal deletions. They will use these analyses to assess the possibility that breast cancer arises through multiple distinct causal pathways and that, in fact, breast cancer represents a heterogenous group of etiologic entities. The investigators point out that studies of this type will help to classify breast cancers into etiologically relevant categories, leading eventually to the identification of specific modifiable risk factors.